1. Field of the Invention
The present invention relates to an image display system that modulates the illumination light transmitted from a light source according to image signals for projecting and displaying images. More particularly, this invention relates to an image display system implemented with a light source controlled to project light of variable intensities to further increase the number of gray scales for improving the quality of image display.
2. Description of the Related Arts
After the dominance of CRT technology in the display industry over the past 100 years, the Flat Panel Display (hereafter FPD) and Projection Display technologies are now gaining popularity because of a smaller form-factor of the display control system while enabled to project and display image of greater size onto a bigger display screen. Among several types of projection display systems, projection display systems using micro-display are gaining consumers' recognition because of high performance of picture quality as well as lower cost than the display systems implemented with FPDs. There are two types of micro-display technologies implemented in the projection display systems now made available in the market. The first type of display system is the micro-LCD (Liquid Crystal Display) system and the other type display system is the display system that implements the micromirror technology. Because a micromirror device uses un-polarized light, a micromirror device has an advantage that the display image projected from the micromirror device has a higher brightness over the display image projected from a micro-LCD system, which uses polarized light.
Even though there are significant advances of the technologies in implementing an electromechanical mirror device as a spatial light modulator (SLM) in recent years, there are still limitations and difficulties when it is employed to provide a high quality image. Specifically, when the images are digitally controlled, the image quality is adversely affected due to the fact that the images are not displayed with sufficient levels of gray scales.
An electromechanical mirror device is drawing a considerable interest for application as a spatial light modulator (SLM). The electromechanical mirror device includes “a mirror array” that has a large number of mirror elements. In general, the mirror elements from 60,000 to several millions are arranged on a surface of a substrate in an electromechanical mirror device. In order to better understand this invention, the following background descriptions are presented followed by discussions of why the conventional technologies as now available on the marketplace and the state of the art of image display devices still confronted with technical difficulties and limitations.
Outline of the Device
The first is a description of a mirror device. Image projection apparatuses implemented with a spatial light modulator (SLM), such as a transmissive liquid crystal, a reflective liquid crystal, a mirror array and other similar image modulation devices, are widely known.
A spatial light modulator is formed as a two-dimensional array of optical elements, ranging from tens of thousands to millions of miniature modulation elements, with individual elements enlarged and displayed as the individual
pixels corresponding to an image to be displayed onto a screen by way of a projection lens.
The spatial light modulators generally used for projection apparatuses primarily include two types, i.e., a liquid crystal device for modulating the polarizing direction of incident light by sealing a liquid crystal between transparent substrates and providing them with a potential, and a mirror
device deflecting miniature micro electro mechanical systems (MEMS) mirrors with electrostatic force and controlling the reflecting direction of illumination light.
One embodiment of the above described mirror device is disclosed in U.S. Pat. No. 4,229,732, in which a drive circuit using MOSFET and deflectable metallic mirrors are formed on a semiconductor wafer substrate. The mirror
deflects to different angles according to the electrostatic force supplied from the drive circuit and is capable of changing the reflecting direction of the incident light.
Meanwhile, U.S. Pat. No. 4,662,746 has disclosed an embodiment in which one or two elastic hinges retain a mirror. If the mirror is retained by one elastic hinge, the elastic hinge functions as bending spring. If the mirror is retained by
two elastic hinges, they function as torsion springs to incline the mirror, thereby the deflecting the reflecting direction of incident light.
As described above, the ON-and-OFF states of a micromirror control scheme as that implemented in U.S. Pat. No. 5,214,420 and by most conventional display systems limits display quality. Specifically, conventional control circuits
limits the gray scale (PWM between ON and OFF states) since it is limited by the LSB (least significant bit, or the least pulse width). Due to the ON-OFF states implemented in conventional systems, there is no way to provide a shorter
pulse width than the LSB. The lowest brightness, which determines the gray scale, is the light reflected during the least pulse width. The limited gray scale leads lower image quality.
Specifically, FIG. 1C exemplifies a conventional control circuit for a micromirror according to U.S. Pat. No. 5,285,407. The control circuit includes memory cell 32. Various transistors are referred to as “M*” where * designates a
transistor number and each transistor is an insulated gate field effect transistor. Transistors M5 and M7 are p-channel transistors; transistors, M6, M8, and M9 are n-channel transistors. The capacitances, C1 and C2, represent the capacitive
loads of the memory cell 32. Memory cell 32 includes an access switch transistor M9 and a latch 32a, which is the basis of the static random access switch memory (SRAM) design. All access transistors M9 in a row receive a DATA signal from
a different bit-line 31a. The particular memory cell 32 is accessed and written by turning on the appropriate row select transistor M9, using the ROW signal functioning as a wordline. Latch 32a is formed from two cross-coupled inverters,
M5/M6 and M7/M8, which permit two stable states. State 1 is Node A high and Node B low and state 2 is Node A low and Node B high.
The mirror driven by a drive electrode abuts a landing electrode structured differently from the drive electrode, thereby a prescribed tilt angle is maintained. A “landing chip”, which possesses a spring property, is formed on the point of contact between the landing electrode and the mirror. This configuration enhances the deflection of the mirror to the reverse direction upon a change in the control. The parts forming the landing chip and the landing electrode are maintained at the same potential so that contact will not cause a shorting or other similar disruption.
Outline of Time Domain Grayscale Control
The following description explains the pulse-width modulation (PWM) control using Time Domain Grayscale Control.
As described above and shown in FIG. 1A, an operation for switching the mirror by the control circuit deflects the micromirrors in either an ON or an OFF angular orientation.
The brightness, i.e., the gray scales of display for a digitally
controlled image system is determined by the length of time the micromirror stays at an ON position. The length of time a micromirror is controlled at an ON position is controlled by a multiple bit word. FIG. 1D shows the “binary time intervals” when the micromirror is controlled by a four-bit word. As in FIG. 1D, the time durations have relative values of 1, 2, 4, 8 that, in turn, define the relative brightness for each of the four bits where 1 is the least significant bit and 8 is the most significant bit. According to the control mechanism as shown, the minimum controllable differences between gray scales for showing different brightness is the brightness represented by the “least significant bit” that can maintain the micromirror at an ON position.
In a simple exemplary display system operated with an n bits brightness control signal for controlling the gray scales, the frame time is divided into 2″−1 equal time slices. For a 5.56 milliseconds frame period for each color (1/(60×3) seconds=5.56 milliseconds) and n-bit intensity values, the time slice is 5.56/(2″−1) milliseconds.
Having established these time slices for controlling the length of time for displaying each pixel in each frame, the pixel intensities are determined by the number of time slices represented by each bit. Specifically, a display of a black pixel is represented by 0 time slices. The intensity level represented by the LSB is 1 time slice, and maximum brightness is 2″−1 time slices. The number time slices that a micromirror is controlled to operate at an On-state in a frame period determines a specifically quantified light intensity of each pixel corresponding to the micromirror reflecting a modulated light to that pixel. Thus, during a frame period, each pixel corresponding to a modulated micromirror controlled by a control word with a quantified value of more than 0 is operated at an on state for the number of time slices that correspond to the quantified value represented by the control word. The viewer's eye integrates the pixels' brightness so that the image appears the same as if it were generated with analog levels of light.
For addressing deformable mirror devices, a pulse width modulator (PWM) receives the data formatted into “bit-planes”. Each bit-plane corresponds to a bit weight of the intensity value. Thus, if each pixel's intensity is represented
by an n-bit value, each frame of data has n bit-planes. Each bit-plane has a 0 or 1 value for each display element. In the example described in the preceding paragraphs, each bit-plane is separately loaded during a frame. The display elements are addressed according to their associated bit-plane values. For example, the bit-plane representing the LSBs of each pixel is displayed for 1 time slice.
As shown in FIG. 3A, the signal bit data described as “Dn” where n is 0 through 5 has 6 bits in total and each bit contains 0 or 1. The bit Dn will be transferred to the memory of the SLM and kept during the time duration of (2″−1) times Bit Time which is the minimum controllable time of the SLM.
As shown in FIG. 3B, the combination of 0 and 1 in the signal data, D0 through D5 with the illumination, the system can deliver output energy from 0 through 63 with the minimum increment of 1. Therefore this system can control the output in 64 levels of light energy and can achieve the grayscale of 64.
While the mirror is always ON with all the incoming data bits are 1 during all the bit durations, the output light energy will be 63 times the energy reflected by a mirror during one Bit Time. Therefore the average output light energy per Bit Time will be 63/63=1 or 100%. This system can deliver all possible light energy received from the light sources.
In spite of the advantage of brightness, the required speed of data transfer is very high. For an example, an 8 bit grayscale is common with HD-TV format which has 1920 horizontal dots with 1080 lines vertically. At 8 bit grayscale, the “Bit Time” will be 1/(60×3×256) seconds or about 20 micro-seconds. This requires the system to transfer 1920×1080 bits per “Bit Time” or about 20 micro-seconds. This equates to about 95.6 Giga bits per second. This speed is not easy to achieve even with the latest technology without costly circuit.
Outline of Light Width Grayscale Control (LWGC)
One embodiment of Light Width Grayscale Control is disclosed in U.S. Pat. No. 5,903,323, wherein the concept of LWGC is explained well.
FIG. 4A shows an example of LWGC, wherein the bit duration for each bit at the SLM is equal and the time width of illumination has the time duration of 1/2^(5−n) of the Bit Time for the data bit Dn. The same characteristics of “Light Width Grayscale Control” method are shown in FIG. 4B. The maximum output of light energy will be obtained when all the signal bits are 1 and the total output energy in a frame will be 1.96875 out of 6 Bit Times. Therefore the average output energy per Bit Time will be 1.96875/6=0.328. This means that the maximum output light energy at all the incoming signal bits of 1, or simply “the brightness” of Light Width Grayscale Control is 32.8% of “Time Domain Grayscale Control” if a same light source is used. The maximum number of brightness levels or the grayscale will be 64 which is the same as that of “Time Domain Grayscale Control”. The grayscale is same as that of “Time Domain Grayscale Control”. On the other hand, the speed required to transfer the signal bits is much less and the Bit Time will be a frame period, 1/(60×3) seconds, divided by 8 or about 694 micro-seconds, if 8 bit grayscale is used. This equates to about 3 Giga bits per second, which is far less than 95.5 Giga bits per second of Time Domain Grayscale Control.
Outline of Light Intensity Grayscale Control (LIGC)
One embodiment of Light Width Grayscale Control is disclosed in U.S. Pat. No. 6,232,963, wherein the concept of LIGC is explained.
“Light Intensity Grayscale Control” method is shown in FIG. 5. The characteristics are almost same as those of “Light Width Grayscale Control”. The brightness of “Light Intensity Grayscale Control” is about 32.8% of “Time Domain Grayscale Control”, if a same light source is used. The grayscale is the same as that of “Time Domain Grayscale Control”. The speed required to transfer data to the SLM is the same as that of “Light Width Grayscale Control”. The advantage of LIGC is the required speed to control the light source is less than that of LWGC. As shown, both LWGC and LIGC are attractive for slower data transfer circuits, but the loss of the brightness is significant and loses as much as ⅔ of the available brightness.
With the above background descriptions, the following discussion further explain the limitations and technical difficulties of the conventional technologies.
Referring to FIG. 1A for an image display system 1 including a screen 2 is disclosed in a reference U.S. Pat. No. 5,214,420. A light source 10 is used for generating light energy for illuminating the screen 2. The generated light 9 is further collimated and directed toward a lens 12 by a mirror 11. Lenses 12, 13 and 14 constitute a beam collimator to collimate light 9 into a column of light 8. A spatial light modulator (SLM) 15 is controlled on the basis of data input by a computer 19 via a bus 18 and selectively redirects the portions of light from a path 7 toward an enlarger lens 5 and onto screen 2. The SLM 15 has a mirror array includes switchable reflective elements 17, 27, 37, and 47 each comprising a mirror 32 connected by a hinge 30 and supported on a surface 16 of a substrate in the electromechanical mirror device as shown in FIG. 1B. When the element 17 is in one position, a portion of the light from the path 7 is redirected along a path 6 to lens 5 where it is enlarged or spread along the path 4 to impinge on the screen 2 so as to form an illuminated pixel 3. When the element 17 is in another position, the light is redirected away from the display screen 2 and hence the pixel 3 is dark.
Most of the conventional image display devices such as the devices disclosed in U.S. Pat. No. 5,214,420 are implemented with a dual-state mirror control that controls the mirrors to operate at a state of either ON or OFF. The quality of an image display is limited due to the limited number of gray scales. Specifically, in a conventional control circuit that applies a PWM (Pulse Width Modulation), the quality of the image is limited by the LSB (least significant bit) or the least pulse width as control related to the ON or OFF state. Since the mirror is controlled to operate in an either ON or OFF state, the conventional image display apparatuses have no way to provide a pulse width to control the mirror that is shorter than the control duration allowable according to the LSB. The least quantity of light, which determines the least amount of adjustable brightness for adjusting the gray scale, is the light reflected during the time duration according to the least pulse width. The limited gray scale due to the LSB limitation leads to a degradation of the quality of the display image.
Specifically, FIG. 1C shows an exemplary control circuit for controlling a mirror element according to the disclosures made in U.S. Pat. No. 5,285,407. The control circuit includes a memory cell 32. Various transistors are referred to as “M*” where “*” designates a transistor number and each transistor is an insulated gate field effect transistor. Transistors M5 and M7 are p-channel transistors; while transistors M6, M8, and M9 are n-channel transistors. The capacitances C1 and C2 represent the capacitive loads in the memory cell 32. The memory cell 32 includes an access switch transistor M9 and a latch 32a, which is based of a Static Random Access Switch Memory (SRAM) design. The transistor M9 connected to a Row-line receives a DATA signal via a Bit-line. The memory cell 32 written data is accessed when the transistor M9, which has received the ROW signal on a Word-line is turned on. The latch 32a includes two cross-coupled inverters, i.e., M5/M6 and M7/M8, which permit two stable states, that is, a state 1 is Node A high and Node B low, and a state 2 is Node A low and Node B high. The control circuit as illustrated in FIG. 1C controls the micromirrors to switch between two states and the control circuit drives the mirror to oscillate to either an ON or OFF deflected angle (or position) as shown in FIG. 1A.
The minimum quantity of light controllable to reflect from each mirror element for image display, i.e., the resolution of gray scale of image display for a digitally controlled image display apparatus, is determined by the least length of time that the mirror is controllable to hold at the ON position. The length of time that each mirror is controlled to hold at an ON position is in turn controlled by multiple bit words. FIG. 1D shows the “binary time periods” in the case of controlling SLM by four-bit words. As shown in FIG. 1D, the time periods have relative values of 1, 2, 4, and 8 that in turn determine the relative quantity of light of each of the four bits, where the “1” is least significant bit (LSB) and the “8” is the most significant bit. According to the PWM control mechanism, the minimum quantity of light that determines the resolution of the gray scale is a brightness controlled by using the “least significant bit” for holding the mirror at an ON position during a shortest controllable length of time.
As illustrated in FIG. 2A, when adjacent image pixels are displayed with a brightness controlled by very coarse gray scales, the adjacent pixels may be displayed with great differences of quantity of light, thus, artifacts are shown between these adjacent image pixels. That leads to the degradations of display image qualities. The degradations of image qualities are specially pronounced in the bright areas of image when there are “bigger gaps” of gray scale, i.e. quantity of light, between adjacent image pixels. For example, the bright areas are generally observed on the forehead, the sides of the nose and the upper arm in an image of human and there are artifacts shown in these bright areas when displayed with gray scales of coarse resolutions. The artifacts are caused by a technical limitation that the digitally controlled image does not obtain sufficient number of the gray scale, i.e. the levels of the quantity of light. At the bright spots of display, e.g., the forehead, the sides of the nose and the upper arm, the adjacent pixels are displayed with visible gaps of light intensities. When the levels of gray scales are increased, the image degradation is significantly reduced even with only twice more levels of gray scales as illustrated in FIG. 2B.
Therefore, when the mirrors are controlled to operate at either ON or OFF position then the adjustable quantity of light of a displayed image is controlled by the length of time each mirror is held at the ON position. In order to increase the number of the levels of the controllable quantity of light, the switching speed of the ON and OFF positions for the mirror must be increased. A higher number of bits is therefore required to control the ON and OFF states of the micromirrors. However, when the switching speed of the mirror deflection is increased, a stronger hinge for supporting the mirror is necessary to sustain a required number of switches of the ON and OFF positions for the mirror deflection. Furthermore, in order to drive the mirrors provided strengthened hinge toward the ON or OFF positions, it becomes necessary to apply a higher voltage to the electrode. The higher voltage may exceed twenty volts and may even be as high as thirty volts. The mirrors produced by applying the CMOS technologies probably is not appropriate for operating the mirror at such a high range of voltages, and therefore the DMOS mirror devices may be required. In order to achieve a control of higher number of the gray scales, a more complicated production process and larger device areas are required to produce the DMOS mirror. Conventional mirror controls are therefore faced with a technical problem that higher level of gray scales and can only be achieved by operating the micromirrors at a range of higher voltage to maintain the benefits of manufacturing a smaller image display apparatus.
There are many patents related to the control of quantity of light. These patents include U.S. Pat. Nos. 5,589,852, 5,617,243, 5,668,611, 5,706,061, 5,903,323, 6,232,963, 6,262,829, 6,592,227, 6,648,476, 6,819,064, and 6,975,366. There are further patents and patent applications related to different sorts of light sources. These patents include U.S. Pat. Nos. 5,442,414, 6,036,318 and Application 20030147052. Also, The U.S. Pat. No. 6,746,123 has disclosed particular polarized light sources for preventing the loss of light. However, these patents or patent applications do not provide an effective solution to attain a sufficient number of the gray scale in the digitally controlled image display system.
Furthermore, there are many patents related to a spatial light modulation that includes the U.S. Pat. Nos. 2,025,143, 2,682,010, 2,681,423, 4,087,810, 4,292,732, 4,405,209, 4,454,541, 4,592,628, 4,767,192, 4,842,396, 4,907,862, 5,214,420, 5,287,096, 5,506,597, and 5,489,952. However, these inventions do not provide a direct solution for a person skilled in the art to overcome the above-discussed limitations and difficulties. Therefore, a need still exists in the art of image display systems applying digital control of a micromirror array as a spatial light modulator to provide new and improved systems such that the above-discussed difficulties can be resolved. The major difficulty that hinders the increase of the gray scales of image display is caused by the ON-OFF control scheme of the conventional systems that limits the minimum ON time to adjust the brightness of a display image. The minimum ON time determines the height of the steps of gray scale in FIG. 2. There is no way to provide the adjustable brightness that is lower than the step by controlling the micromirrors. In order to overcome the problems of the degradation of picture quality it is necessary to increase the level of adjustable brightness with adjustable brightness lower than the step shown in FIG. 2.
There is an increasing demand for an image display system to display image with higher image quality. One of the determining factors for displaying an image with improved image quality is to increase the pixel resolution such as in a high definition (HD) display system a HD-level (1920×1080) high resolution is gaining popularity. However, another important factor for improving the image quality is to increase the level of gray scales and as of now most of image display systems still use an 8-bit display mode (8 bits for each of RGB colors, for example). In order to improve the image display quality, it is foreseeable that in the near future, it is desirable and necessary that an image display system uses a greater number of gray scales, for example, gray scales controlled by digital word 10 bits or more.
A micromirror device is implemented as a display system for wide screen image display configured either as a front projector or a rear-projection TV. A micro-mirror device as now implemented in the display systems deflects the illumination light in two directions, ON and OFF. Several types of grayscale control methods are proposed as illustrated in FIG. 3A, FIG. 4A and FIG. 5. The most commonly used PWM is illustrated in FIG. 3A, wherein the time duration of the signal bit (D0 through D5) is proportional to 2″ and the intensity of illumination stays same. Because some other grayscale control methods will be discussed here, the conventional PWM described above will be referred to as “Time Domain Grayscale Control”. Another proposed method is “Light Width Grayscale Control” by varying the pulse width of light and keeping the time duration of signal bits constant and shown in FIG. 4A. The third method proposed is “Light Intensity Grayscale Control” by varying the intensity of light and keeping the time duration of signal bits constant and shown in FIG. 5.
Each method has advantages and disadvantages. The important characteristics of “Time Domain Grayscale Control” are shown in FIG. 3B. The time duration of a frame is 63 times the “minimum controllable time” of SLM (hereafter “Bit Time”). The maximum output of light energy will be obtained when all the signal bits are 1 and the total output energy in a frame will be 63 units of the light intensity of 1 multiplied by Bit Duration as shown in FIG. 3B. Therefore the average output energy per Bit Time will be 63/63=1. The maximum number of brightness levels will be 64 including 0 through 63, in other words, the grayscale is 64. This method is required to send at least one bit of signal to each of pixel in the entire pixel array in one Bit Time (the minimum controllable time of SLM). For example, nowadays 8 bit grayscale is common with HD-TV format which has 1920 horizontal dots with 1080 lines vertically. At 8 bit grayscale, the “Bit Time” will be 1/(60×3×256) seconds or about 20 micro-seconds. This requires the system to transfer 1920×1080 bits per “Bit Time” or about 20 micro-seconds. This equates to about 95.6 Giga bits per second. This speed is not easy to achieve even with the latest technology without costly circuit.
The same characteristics of “Light Width Grayscale Control” method are shown in FIG. 4B. The maximum output of light energy will be obtained when all the signal bits are 1 and the total output energy in a frame will be 1.96875 out of 6 Bit Times. Therefore the average output energy per Bit Time will be 1.96875/6=0.328. This means that the maximum output light energy at all the incoming signal bits of 1, or simply “the brightness” of Light Width Grayscale Control is 32.8% of “Time Domain Grayscale Control” if a same light source is used. The maximum number of brightness levels or the grayscale will be 64 which is the same as that of “Time Domain Grayscale Control”. The grayscale is same as that of “Time Domain Grayscale Control”. On the other hand, the speed required to transfer the signal bits is much less and the Bit Time will be a frame period, 1/(60×3) seconds, divided by 8 or about 694 micro-seconds, if 8 bit grayscale is used. This equates to about 3 Giga bits per second, which is far less than 95.5 Giga bits per second of Time Domain Grayscale Control.
“Light Intensity Grayscale Control” method is shown in FIG. 5. The characteristics are almost same as those of “Light Width Grayscale Control”. The brightness of “Light Intensity Grayscale Control” is about 32.8% or less of “Time Domain Grayscale Control”, if a same light source is used. The grayscale is same as that of “Time Domain Grayscale Control”. The speed required to transfer data is same as that of “Light Width Grayscale Control”.
With these comparisons among the three methods, obviously “Time Domain Grayscale Control” has an advantage in brightness and it is about 3 times brighter than the other two methods, if a same light source is used. On the other hand, it requires substantially higher data transfer speed than the other two methods. The consumers obviously chose brighter TVs rather than inexpensive TVs and Time Domain Grayscale Control became an industry standard.
However, the present technology does not allow inexpensively higher level of grayscale using Time Domain Grayscale Control than 8 bits. In spite of the technological limitation, the market is moving toward much higher level of grayscale. The grayscale of DVD is 8 bits, but newly introduced “Blue-ray” disc has 16 bit grayscale which provides much smoother pictures. There is a need for substantially higher level of grayscale which can be achieved in an inexpensive way.